Large enterprises have many modeled processes and potentially many instances of each process. Sub-process patterns that are common across multiple processes in the enterprise may exist but may not be identified as such. Furthermore, while the similar sub-process patterns may exist in multiple processes, they may not have the same performance measure or efficiency. For instance, a sub-process pattern that works well in one process model may not perform as well in another process model. More concretely, the same set of tasks executed in a first project that worked well in completing that first project may not work as well in a different second project. Regardless, without knowing the difference in the performance results, the same set of tasks may be used for the different second project. Thus, applicants in the present disclosure have recognized that identifying similar patterns and also determining their performance measures in different projects or process model contexts would be valuable.